estimation of particulate organic carbon flux in relation to photosynthetic production in a shallow...
TRANSCRIPT
Estimation of particulate organic carbon flux in relationto photosynthetic production in a shallow coastal area
in the Seto Inland Sea
Hitomi Yamaguchi a, Shigeru Montani a,*, Hiroaki Tsutsumi b,Ken-ichiro Hamada c, Naoko Ueda c
a Graduate School of Fisheries Sciences, Hokkaido University, Minato 3-1-1, Hakodate, Hokkaido 041-8611, Japanb Prefectural University of Kumamoto, Tsukide, Kumamoto 862-0920, Japan
c Kitakyusyu City Instituite of Environmental Sciences, Tobata, Kitakyusyu 804-0082, Japan
Accepted 15 October 2002
Abstract
Sediment trap experiments were carried out three times from 1999 to 2000, in the western part of the Seto Inland Sea (Suo-
Sound), Japan. We investigated both the particulate flux and the composition of chemical substances in the sediment trap samples.
Based on the results, we discuss the origin of particulate organic carbon (POC) collected by the sediment traps in a coastal area.
Moreover, we purposed to estimate the flux of the portion of the POC that is derived from phytoplankton photosynthesis. The fluxes
of POC varied between 677 and 3424 mgCm�2 d�1. Significant positive correlations between POC and aluminum (Al) fluxes
suggested that these components show almost the same behaviour. The mean value of the Al flux was about eight times higher than
that of Al burial rates on the sediment surface. Therefore, it seems that the POC flux observed with the sediment traps was con-
siderably overestimated. Moreover, judging from the fact that Al is a typical terriginous element, it seems that most of the POC
collected in the sediment traps derived from the re-suspended surface sediment or sediment transported laterally from shallow flanks
such as intertidal mudflats. The fluxes of chlorophyll a (Chl a) were independent of the POC fluxes, and a relatively consistent
correlation was found between Chl a abundance in the water column and the Chl a flux. Moreover, surface sediment Chl a con-tent was approximately 100 times lower than that of suspended matter. Therefore, resuspension and terriginous contributions to
Chl a collected in sediment traps are likely to be negligible. The POC content in the trap samples varied between 22.4 and 70.7
mg g�1 dry weight. The variations of POC contents were positively correlated with the Chl a contents: POCðmgg�1Þ ¼ 76:5�Chl aðmgg�1Þ þ 26:0 (r ¼ 0:95, p < 0:01, n ¼ 9). This result shows that POC contents strongly corresponded with phytoplankton
and their debris. It was also considered that the fraction of POC derived from phytoplankton primary production could be esti-
mated as Chl a content times a certain factor. In this study, we estimated the flux of the portion of the POC originating from
phytoplankton production by multiplying the Chl a fluxes by 76.5 (the mean POC:Chl a ratio in the trap samples). These valuesvaried between 308 and 758 mgCm�2 d�1, and accounted for 35:1� 21:2% of total POC flux. Although the amount of POC that
originates from phytoplankton photosynthesis was a small portion of total POC flux, it seems to be a large portion of potential
primary production in the water column.
� 2003 Elsevier Science Ltd. All rights reserved.
Keywords: Sediment trap; POC flux; Coastal zone; Phytoplankton; Biological production
1. Introduction
Phytoplankton photosynthesis is one of the majorsources of particulate organic matter (POM), since the
products that ultimately depend on phytoplankton
photosynthesis production include not just phyto-
plankton itself, but also its debris and zooplankton fecal
pellets. A considerable part of POM produced by phy-toplankton is lost through sinking processes (Smetacek,
1985; Wassmann, 1990), and this POM is the major
source of nutritional input to benthic organisms (Sme-
tacek, 1984; Graf, 1992). Therefore, a quantitative esti-
mation of the downward flux of POM originating from
*Corresponding author. Tel./fax: +81-138-40-8871.
E-mail address: [email protected] (S. Montani).
0025-326X/03/$ - see front matter � 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0025-326X(02)00414-9
www.elsevier.com/locate/marpolbul
Marine Pollution Bulletin 47 (2003) 18–24
phytoplankton photosynthesis is useful information for
understanding the coupling between pelagic and benthic
ecosystems (Smetacek et al., 1978; Gardner et al., 1989;
Fitzgerald and Gardner, 1993).In estimating the downward flux, many researchers
use sediment traps (Blomqvist and Hakanson, 1981).
Sediment traps have commonly been used to reveal
various phenomena occurring in the sea, and many
valuable results have been obtained with them. How-
ever, downward flux measurements using sediment traps
have some technical problems (e.g. Gardner, 1980;
Tsunogai et al., 1980; Blomqvist and Kofoed, 1981).One of them is the difficulty of determining the origins
of particles (Nakanishi et al., 1992; Roden et al., 1995),
which makes it difficult to confirm whether or not POM
flux reflects primary production. This problem is very
serious in terms of understanding the coupling between
pelagic and benthic ecosystems. Particularly, coastal
zones are shallow and feature high turbidity due to al-
lochthonous matter in addition to phytoplankton andsubstances from their decomposition. Therefore, it
seems that POM flux does not always reflect the phyto-
plankton primary production. Some methods of cor-
recting for the interface portion in settling POM have
been proposed (e.g. Gasith, 1975; Noriki et al., 1985;
Larsson et al., 1986; Roden et al., 1995; Honda et al.,
2000), but alternatives remain limited.
In this study, sediment trap experiments were con-
ducted in a coastal zone. We investigated both theparticulate fluxes and the composition of chemical
substances obtained in the sediment traps. Based on
these results, we discuss the origin of particulate organic
carbon (POC) collected by sediment traps. In addition,
we tried to estimate the flux of the portion of the POC
that is derived from the primary production of phyto-
plankton.
2. Materials and methods
2.1. Study area
The sampling stations were located in Suo-Sound, in
the western part of the Seto Inland Sea. Topographically,
the study area is a gently sloping plain, containing in-tertidal mudflats (Fig. 1). Although no large rivers run
into the study area, the particulate suspended matter
(SPM) concentration is more than 10 mg l�1 (Yamagu-chi, 2001). Therefore, it seems that re-suspended particles
play a major role in controlling the SPM concentration
Fig. 1. Location map of sampling stations in the western part of Suo-sound.
H. Yamaguchi et al. / Marine Pollution Bulletin 47 (2003) 18–24 19
of the water column. Strong interactions between the
phytoplankton primary production and benthic animal
secondary production in Suo-Sound have been sug-
gested. Nevertheless, little is still known about the POCdownward flux in relation to phytoplankton photosyn-
thesis (Montani et al., 2002).
2.2. Sampling and sample processing
Sediment trap experiments were conducted on 21–22
April and 7–8 October of 1999, and 19–20 April of 2000.
In this study, we used the M-type sediment traps(Montani et al., 1988). Sediment traps were deployed
mainly at 3 m above the bottom, and left for only 24-h
periods to minimize decomposition effects (Taguchi,
1982; Clavier et al., 1995). The entire contents of re-
trieved sediment traps including particles were filtered
using glass fiber filters (pre-combusted Whatman GF/F
filter). These filters were also used for the analysis of
POC, aluminum (Al) and chlorophyll a (Chl a). Whenthe sediment traps were deployed, both SPM samples
and bottom sediment samples were also collected. The
SPM samples were collected at intervals of 2 m from the
surface to the bottom with a Van Dorn bottle water
sampler. They were concentrated by pressure filtration,
the same as sediment trap samples, and used for the
determination of POC and Chl a contents. Bottom
sediment samples were collected with a gravity corer.The surface sediment (upper 1 cm) was sliced with a
plastic plate, and used for the analysis of total organic
carbon (TOC), Al and Chl a. The Chl a content
was determined using a spectrophotometric method
(Lorenzen, 1967). Subsamples of the freeze-dried mate-
rial were used to measure the organic carbon (both POC
and TOC) and Al. Organic carbon content was deter-
mined by high temperature oxidation using a CHNanalyzer (Yanaco MT-3 or Fisions NC-1500) after car-
bonates had been removed with hydrochloric acid.
Sediment digested by hot hydrochloric acid was used for
the analysis of Al. The content of Al was determined
with a modified spectrophotometric method using oxine
(Sandell, 1959).
3. Results and discussion
3.1. Observed chemical fluxes
The fluxes measured in the experimental periods are
shown in Table 1. The highest and lowest POC valueswere 677 and 3424 mgCm�2 d�1, respectively. Both ofthese values were found in October 1999 at Stn. D, al-
though the difference in depth at which the two samples
were taken was only 4 m (Table 1). The results of cor-
relation analysis between POC flux, total mass flux, Al
flux and Chl a flux are shown in Table 2. Significant
positive correlations were found between POC flux, total
mass flux and Al flux. These results suggest that thesecomponents have almost the same behaviour. On the
other hand, the Chl a flux was independent of the otherfluxes (Table 2). For instance, Chl a flux showed the
lowest value in April 2000 at Stn. A, but none of the
other components were present at low values (Table 1).
To examine the reasons for this difference, we investi-
gated the relationship between the concentrations of
respective components and their fluxes in the watercolumn (Fig. 2). The concentration of Chl a in the water
Table 1
Observed particulate flux and content of chemical components in trap samples
Date Station depth (m) Fluxes Contents
Total mass POC Al Chl a POC Al Chl a
(gm�2 day�1) (mgm�2 day�1) (mg g�1 dry weight)
April 1999 A 6 89.0 1992 4253 5.0 22.4 47.8 0.056
October 1999 A 6 26.7 1245 1238 8.0 46.5 46.3 0.297
B 4 54.7 1830 2563 6.2 33.5 46.9 0.114
C 4 14.7 996 534 7.3 67.8 36.4 0.500
6 24.6 1439 1190 9.9 58.3 48.3 0.402
D 6 9.6 677 288 6.0 70.7 30.1 0.625
10 92.0 3424 4699 7.3 37.2 51.1 0.080
April 2000 A 6 90.5 2736 4717 4.0 30.2 52.1 0.044
B 3 70.3 2486 2923 5.7 35.4 41.6 0.081
Average 52.5 1870 2490 6.6 44.7 44.5 0.244
(�S.D.) 34.2 889 1770 1.8 17.2 7.2 0.219
Table 2
Correlation coefficients between observed particulate flux (n ¼ 9).
p-values: �p < 0:01, all other p > 0:05
Flux Total mass POC Al Chl a
Total mass 0:91� 0:99� )0.57POC 0:92� )0.32Al 0.50
Chl a
20 H. Yamaguchi et al. / Marine Pollution Bulletin 47 (2003) 18–24
column indicated a significant amount of variance in
Chl a flux (Fig. 2(c)). In contrast, both the POC and the
SPM concentrations were unrelated to their fluxes (Fig.
2(a) and (b)). Therefore, it seemed that their originswere different from that of Chl a flux. To determine thefactors responsible for the POC flux, we focused on
the Al flux. The fluxes of Al varied between 288 and
4717 mgAlm�2 d�1, with a mean value of 2489
mgAlm�2 d�1 (Table 1). The amount of sediment ac-cumulation observed near our study area was 0.21 g m�2
d�1 (Hoshika and Shiozawa, 1985), and the mean Al
content of surface sediment at the four stations in thisstudy was 52:7� 3:8 mg g�1 (n ¼ 7, data was not
shown). Therefore, the burial rate of Al sediment was
estimated to be 303 mgAlm�2 d�1. Aluminum has a
short oceanic residence time (100–200 years) owing toits affinity to particles (Orians and Bruland, 1985), and it
is possible that there is little regeneration of this element
to the water column. Therefore, the net flux of Al can be
regarded as nearly equal to the burial rates onto thesediment surface (Noriki et al., 1985). This indicates that
the Al flux observed with sediment traps does not cor-
respond to the net Al flux. The same suggestion applies
to the POC flux, which was significantly correlated with
Al flux. Judging from the fact that Al is a typical ter-
rigenous element (Brewer et al., 1980), most of the POC
collected by the sediment traps probably originated from
the re-suspended surface sediment or sediment trans-ported laterally from shallow flanks such as intertidal
mudflats. Therefore, the POC that is derived from phy-
toplankton primary production would be a minor com-
ponent during periods of high POC flux in the study area.
3.2. Chemical composition of trap samples
The POC content in the trap samples varied between22.4 and 70.7 mg g�1 dry weight (Table 1). The highestvalue was about 3.2 times larger than the lowest. The
Chl a content varied between 0.044 and 0.625 mg g�1
and underwent large changes (about 14 times) in com-
position during the experimental periods. In contrast,
the Al content of the trap samples fluctuated in a narrow
range between 30.1 and 52.1 mg g�1. Based on the datafrom chemical contents, it is possible to deduce thefactors affecting the POC content of the trap samples. In
Fig. 3, POC contents are shown as a function of Chl acontents. There was a positive correlation between these
two variables, expressed by the following equation:
POCðmgg�1Þ ¼ 76:5Chl aðmgg�1Þþ 26:0ðr ¼ 0:95; p < 0:01; n ¼ 9Þ
This equation suggests that the content of POC is
roughly divided into two fractions. One fraction of thePOC can be estimated by multiplying the measured Chl
Fig. 2. Relationships between the integrated values of individual
standing stocks in the overlying water column and the observed fluxes.
(a) SPM, (b) POC and (c) Chl a.
H. Yamaguchi et al. / Marine Pollution Bulletin 47 (2003) 18–24 21
a contents by a regression coefficient (POCchl). Fig. 3
shows that a certain portion of POC collected by sedi-
ment traps was strongly correlated to Chl a, in otherwords, phytoplankton primary production. Riemann
et al. (1989) indicated that the estimation of living
phytoplankton carbon biomass from Chl a data in eu-trophic environments such as coastal seas might require
a conversion factor between 27 and 67. Judging from the
fact that the present regression coefficient was a 76.5, it
seems phytoplankton itself was an additional contribu-
tor to POCchl. This could indicate that the principalforms of the POCchl were aggregates, recently ingested
phytoplankton, fecal pellets, and the like. On the other
hand, the other fraction of the POC is expressed as y
intercept (POCy). This fraction has a stable content of
POC (26.0 mg g�1). This feature is the same as surfacesediment (Fig. 4), and may have been due to selective
degradation of organic compounds containing low levelsof carbon (e.g. lipids and carbohydrates). Therefore, it
seems that the nature of the majority of POCy is re-
fractory. The POCy may almost entirely reflect the POC
flux, because the POC flux and the Chl a flux were in-dependent of one another (Table 2). From these results,
we judged that the Chl a content could act as a markerin estimating the POC content that originates from
phytoplankton production.
3.3. Origin of Chl a collected by sediment trap
The mean content of Chl a in SPM (634� 385 lg g�1)was 126-fold greater than that in surface sediments
(5:03� 4:74 lg g�1) through the experimental periods
(Fig. 4(a)). However, in the case of organic carbon, the
difference was only 5-fold (Fig. 4(b)). If the collected
matter in the sediment trap was based on both the SPMand surface sediment, the content of any chemical
property in the trapped material could be expressed as a
combination of SPM and surface sediment values
(Gasith, 1975; Taguchi, 1982; Clavier et al., 1995;
Bhaskar et al., 2000):
CTRP ¼ XCSPM þ ð1� X ÞCSDMwhere CTRP is the chemical composition of the materialin the trap, CSPM is the chemical composition of the
SPM in the water column, CSDM is the chemical com-
position of the surface sediment, and X � 100 ð%Þ is therelative contribution of the trapped matter. Therefore, itcan be seen that the Chl a collected by sediment trap hasa smaller impact than organic carbon on the re-sus-
Fig. 3. Correlation between POC content and Chl a content in trapsamples.
Fig. 4. Mean value of (a) Chl a and (b) organic carbon content in the SPM sample and surface sediment sample. Bars indicate standard deviations.
22 H. Yamaguchi et al. / Marine Pollution Bulletin 47 (2003) 18–24
pended fraction. Roden et al. (1995) pointed out the fact
that surface sediment Chl a content is 10–100 times
lower than that of suspended matter, owing to the rapid
decomposition (Sun et al., 1991). Our results re-con-firmed the findings of Roden et al. (1995) that resus-
pension contributions to the Chl a collected in traps arelikely to be negligible.
3.4. Estimation of the POC flux with relation to photo-
synthetic production
From the above findings, we attempted to estimatethe flux of POC with relation to photosynthetic pro-
duction, based on the following. (1) A relatively con-
sistent relationship was found between Chl a abundancein the water column and Chl a flux (Fig. 2(c)). (2) Thecontent of Chl a in the sediment trap sample reflects
the content of POC that is derived from primary pro-
duction in the water column (Fig. 3). (3) Chl a collectedby sediment traps seems to be less affected by the re-suspended fraction (Fig. 4(a)). Therefore, we estimated
the flux of the portion of the POC that originates from
phytoplankton photosynthesis after multiplying the
gross Chl a flux by 76.5 (see Fig. 3). The estimated POCflux varied between 308 and 758 mgCm�2 d�1 (Table 3).The average contribution of the flux of POC that orig-
inates from phytoplankton photosynthesis comprised
approximately 35% of the total POC flux. Using thesimplest method, primary production may be calculated
from both Chl a concentrations in the water column andassimilation number (mgCmgChl a�1 d�1). Tada et al.(1998) reported the phytoplankton productivity in the
Seto Inland Sea (15.5 mgCmgChl a�1 d�1 in summer
and 15.7 mgCmgChl a�1 d�1 in autumn). From these
data, we preliminarily estimated the primary production
in the sampling periods, and deduced the fraction ofprimary production lost to sinking processes (export
ratio). The results showed that the export ratio near the
bottom floor ranges from 38% to 114% (Table 3). This
indicates that the flux of the POC originating from
primary production accounted for a minor portion of
total POC flux; however, it was the major portion of
primary production in the water column.
4. Conclusion
Whether or not POM flux reflects phytoplankton
photosynthesis can be judged from the correlation be-
tween POM fluxes and plant pigment fluxes. If no cor-
relation is found, the fraction of POM that is derived
from phytoplankton photosynthetic production may be
extracted using the relationship between plant pigmentand organic matter contents of the trapped matter. In
estimating the flux of the portion of the POM that is
derived from phytoplankton photosynthesis, we suggest
multiplying the plant pigment flux by the POM:pigment
ratio (e.g. C: Chl a ratio) in the trap sample.
Acknowledgements
We are indebted to Dr. K. Tada, Ms. Y. Eto, Ms. M.
Fukumoto and all other group members for assistance
in the field. We also thank Dr. O. Oku for very helpful
comments on the analysis.
References
Bhaskar, P.V., Gardozo, E., Giriyan, A., Garg, A., Bhosle, N.B., 2000.
Sedimentation of particulate matter in the Dona Paula Bay, west
coast of India during November to May 1995–1997. Estuaries 23,
722–734.
Blomqvist, S., Hakanson, L., 1981. A review on sediment traps in
aquatic environments. Archiv f€uur Hydrobiologie 91, 101–132.Blomqvist, S., Kofoed, C., 1981. Sediment trapping––A sub-
aquatic in situ experiment. Limnology and Oceanography 26, 585–
590.
Brewer, P.G., Nozaki, Y., Spencer, D.W., Fleer, A.P., 1980. Sediment
trap experiments in the deep North Atlantic: isotopic and elemental
fluxes. Journal of Marine Research 38, 703–728.
Clavier, P.J., Chardy, P., Chevillon, C., 1995. Sedimentation of
particulate matter in the south-west lagoon of New Caledonia:
Table 3
The flux of the portion of the POC that originates from primary production. Primary production rates that were preliminarily estimated and export
ratios are also shown
Date Station Depth (m) Estimated POC flux�
(mgm�2 day�1)
Contribution to total
POC flux (%)
Estimated primary pro-
duction (mgm�2 day�1)
Export ratio (%)
April 1999 A 6 379 19.0 415 91.3
October 1999 A 6 608 48.8 987 61.6
B 4 476 26.0 1014 46.9
C 4 562 56.4 – –
6 758 52.7 995 76.2
D 6 457 67.5 – –
10 560 16.4 981 57.1
April 2000 A 6 308 11.3 807 38.2
B 3 438 17.6 385 114
�The flux of the portion of the POC that originates from photosynthetic production.
H. Yamaguchi et al. / Marine Pollution Bulletin 47 (2003) 18–24 23
spatial and temporal patterns. Estuarine, Coastal and Shelf Science
40, 281–294.
Fitzgerald, A.S., Gardner, W.S., 1993. An algal carbon budget for
benthic–pelagic coupling in Lake Michigan. Limnology and
Oceanography 38, 547–560.
Gardner, W.D., 1980. Field assessment of sediment traps. Journal of
Marine Research 38, 41–52.
Gardner, W.S., Eadie, B.J., Chandler, J.F., Parrish, C.C., Malczyk,
J.M., 1989. Mass flux and ‘‘nutritional composition’’ of settling
epilimnetic particles in Lake Michigan. Canadian Journal of
Fisheries and Aquatic Sciences 46, 1118–1124.
Gasith, A., 1975. Tripton sedimentation in eutrophic lakes––simple
correction for the resuspended matter. Verhandlungen Interna-
tional Vereinigung f€uur Theoretische und Angewandte Limnolgie
19, 116–122.
Graf, G., 1992. Benthic–pelagic coupling: a benthic view. Oceanogra-
phy and Marine Biology Annual Review 30, 149–190.
Honda, M.C., Kusakabe, M., Nakabayashi, S., Katagiri, M., 2000.
Radiocarbon of sediment trap samples from the Okinawa trough:
lateral transport of 14C-poor sediment from the continental slope.
Marine Chemistry 68, 231–247.
Hoshika, A., Shiozawa, T., 1985. Sedimenation rates and heavy metal
pollution of sediments in the Seto Inland sea, part 4. Suo-Nada.
Journal of the Oceanographical Society of Japan 41, 283–290.
Larsson, U., Blomqvist, S., Abrahamsson, B., 1986. A new sediment
trap system. Marine Ecology Progress Series 31, 205–207.
Lorenzen, C.J., 1967. Determination of chlorophyll and pheo-pig-
ments: spectrophotometric equations. Limnology and Oceanogra-
phy 12, 343–346.
Montani, S., Tada, K., Okaichi, T., 1988. Purine and pyrimidine bases
in marine particles in the Seto Inland Sea, Japan. Marine
Chemistry 25, 359–371.
Montani, S., Yamaguchi, H., Tsutsumi, H., Hamada, K., Ueda, N.,
2002. Spatial distribution of phosphorous and biological produc-
tion of Suo-Sound (western part of the Seto Inland Sea, Japan) in
summer. Bulletin on Coastal Oceanography 39, 163–171 (in
Japanese with English abstract).
Nakanishi, M., Tezuka, Y., Narita, T., Mitamura, O., Kawabata, K.,
Nakano, S., 1992. Phytoplankton primary production and its fate
in a pelagic area of Lake Biwa. Archiv f€uur Hydrobiologie Beihefte
Ergebnisse der Limnologie 35, 47–67.
Noriki, S., Ishimori, N., Tsunogai, S., 1985. Regeneration of chemical
elements from settling particles collected by sediment trap in Funka
Bay, Japan. Journal of the Oceanographical Society of Japan 41,
113–120.
Orians, K.J., Bruland, K.W., 1985. Dissolved aluminum in the central
North Pacific. Nature 316, 427–429.
Riemann, B., Simonsen, P., Stwnsgaard, L., 1989. The carbon and
chlorophyll content of phytoplankton from various nutrient
regimes. Journal of Plankton Research 11, 1037–1045.
Roden, E.E., Tuttle, J.H., Boynton, W.R., Kemp, W.M., 1995.
Carbon cycling in mesohaline Chesapeake Bay sediments 1: POC
deposition rates and mineralization pathways. Journal of Marine
Research 41, 113–120.
Sandell, E.B., 1959. Colorimetric determination of trace metals, third
ed. Interscience Publishers, New York, 1032pp.
Smetacek, V., 1984. The supply of food to the benthos. In: Fasham,
M.J.R. (Ed.), Flows of Energy and Materials of Marine Ecosys-
tems. Plenum Press, New York, pp. 517–547.
Smetacek, V., 1985. Role of sinking diatom life-history cycles:
ecological, evolutionary and geological significance. Marine Biol-
ogy 84, 239–251.
Smetacek, V., von Brockel, K., Zeitzschel, B., Zenk, W., 1978.
Sedimentation of particulate matter during a phytoplankton spring
bloom in relation to the hydrographical regime. Marine Biology 47,
211–226.
Sun, M., Aller, R.C., Lee, C., 1991. Early diagenesis of chlorophyll-ain long island sound sediments: a measure of carbon flux and
particle reworking. Journal of Marine Research 49, 379–401.
Tada, K., Monaka, K., Morishita, M., Hashimoto, T., 1998. Standing
stock and production rates of phytoplankton and abundance of
bacteria in the Seto Inland Sea, Japan. Journal of Oceanography
54, 285–295.
Taguchi, S., 1982. Sedimentation of newly produced particulate
organic matter in a subtropical inlet, Kaneohe Bay, Hawaii.
Estuarine, Coastal and Shelf Science 14, 533–544.
Tsunogai, S., Uematsu, M., Tanaka, N., Harada, K., 1980. A sediment
trap experiment in Funka Bay, Japan: ‘‘upward flux’’ of particulate
matter in seawater. Marine Chemistry 9, 321–334.
Wassmann, P., 1990. Relationship between primary and export
production in the boreal coastal zone of the North Atlantic.
Limnology and Oceanography 35, 464–471.
Yamaguchi, H., 2001. A study on the biogeochemical cycle of biophilic
elements in Suo Sound, Western part of the Seto Inland Sea, Japan.
MSc thesis, Department of Life Sciences, Kagawa University,
Japan, pp. 1–92.
24 H. Yamaguchi et al. / Marine Pollution Bulletin 47 (2003) 18–24